Adeno-associated virus (AAV) is unique in that it establishes latency by integrating its genome into human chromosome 19 at 19q13.4, termed AAVS1. While this feature has been contemplated as useful for AAV- based strategies for targeted gene delivery, little is known about AAV and its integration mechanism in vivo. Most of the viral life cycle characteristics are concluded from observations in cell culture, as no suitable animal model has been available yet. Recently, we demonstrated that the chromosomal signals required for human site-specific integration are conserved in the mouse genome in a region corresponding to the human target site. Based on this finding we propose to study AAV-mediated transgene targeting in mouse ES cells (Aim 1). This discovery has also opened the opportunity to investigate the biology of wt AAV integration in an as yet unexplored model system. Once ES gene targeting has been established, it will be possible to extend those studies to wild type AAV integration (Aim 1). The ES system will allow us to ask a range of new questions. Does wild type AAV integration affect the differentiation of ES cells into any or all of the lineages? This question is significant since it has been proposed that AAV can infect in utero. Furthermore, introduction of ES cells, that contain integrated AAV, into the blastocyst and subsequent derivation of transgenic mice (in effect knock-out mice for the integration locus) will tell us unambiguously, what - if any - effects can be expected from heterozygous disruption of the integration locus (Aim 2). The need for safe and efficient gene targeting of ES cells has grown since recent developments in stem cell biology have focused considerable attention on the use of cell-based therapies for the treatment of complex diseases. The success of such an approach, however, will require the ability to genetically modify stem cells ex vivo. For example, an obstacle to cell based therapies is the likelihood for the destruction of the newly transplanted cells by immunoreactive cells and/or allo-rejection of the transplanted cells. Therefore, we deem it necessary to investigate whether AAV-based targeted insertion of a transgene into mouse ES cells diminishes concerns about insertional mutagenesis. While in differentiated cells the potential consequences of insertional mutagenesis are apparently negligible, in fast-dividing ES cells, lacking a G1 checkpoint, this concern needs to be addressed. Differentiation assays of ES cells, both in vitro (embryoid body system) and in vivo (transgenic mice), offer the possibility to investigate whether AAVS1 represents a safe targeting site. In addition, these assays will allow us to study if this chromosomal context permits optimal transgene expression in various lineages and tissues. Once established, this system will allow us to test a variety of lineage specific as well as regulated promoters for their activities within this particular chromosomal environment (Aim 3). [unreadable] [unreadable]